We Hold Complaint Data on 2,604 Models. Crash Tests on 311.

Open any car specification page and it reads as finished. Safety rating, fuel economy, recall history, cargo volume — the categories sit in a row, and the ones with nothing behind them are quietly left out of the layout. You never see the hole, so you never ask how big it is.

We track 2,604 models. Here is how many of them carry each kind of fact.

The pattern is who collects it, not what it is

Sort that table by how the number gets made and it stops looking random.

Complaints cover everything because nobody has to go and measure them. A driver files one after the transmission fails, and the record exists. There is no budget, no waiting list, no selection — which is exactly why coverage hits every model we track.

Recalls behave the same way, one step removed. A manufacturer issues one, the notice is published, and it lands against whatever model it lands against. Just over half the catalogue has at least one.

Then the line falls off, and it falls off precisely where somebody has to buy a car and do something to it. Fuel economy needs a lab and a dynamometer. Crash results need a car destroyed on purpose, at roughly forty thousand dollars a run. Coverage there sits at 11.9%.

287 models have the full set

The more useful question is not how many models have crash results, but how many have everything at once — because that is the model you can actually research before you buy.

So a full picture exists for about one model in nine. For 1,038 others — two of every five models we track — the only thing on record is what broke for the people who already own one.

That is not nothing. Complaint patterns are the reason we can tell you which model year of a car to avoid. But it is a different kind of evidence from a crash rating, and a page that shows both in the same grey box invites you to read them as equal.

Why a blank is not a verdict

The temptation with a missing rating is to fill it in yourself. No crash result, so the car must be untested for a reason. Something must be wrong.

Nothing is wrong. Testing programmes pick their subjects by sales volume and by what the public is likely to look up. A model that sold in small numbers can go through its entire production run without a single agency buying one to destroy, and that says nothing whatsoever about how it performs in a collision.

The same logic applies to fuel economy figures that stop partway through a model's life. Certification happens per engine, per model year. Change the engine mid-generation and the older years keep their numbers while the newer ones wait for a fresh test — or never get one, if the volume does not justify it.

What we do with the gaps

Two rules, and they are the reason this article exists at all.

A blank stays blank. We do not carry a figure across from the previous model year, we do not average two trims into one, and we do not borrow a rating from a sibling model that shares a platform. Every one of those tricks would make our pages look more complete and make them less true.

And a number always says where it came from. When you see fuel economy on one of our pages, it came from a certification record for that engine in that year. When you see a crash rating, a specific agency produced it. When you see neither, the honest answer is that nobody has published one — and you should weigh the car on the evidence that does exist.

Running a free VIN check works the same way. It reads the seventeen characters, returns what the record holds for that exact vehicle, and stops there.